Simultaneous spectrum speciation
As a measure of organic compounds, Fourier transform infrared (FTIR) spectrometry has its origins firmly rooted in laboratory analysis. Cameron Stathers and Colin Blackmore, ETI Group Ltd, consider its transferral to the industrial environment.Today, FTIR instruments have been made much more reliable, robust and transportable, providing an alternative to other multi-component techniques such as GC (pictured), MS or non-dispersive IR.
Complex gas mixture
The various techniques for monitoring organic compounds fall into two main categories: total organic or VOC monitoring; and those systems which are able to separate, or "speciate", individual organic compounds in a complex gas mixture. In some instances, such as that of gas chromatography (GC), the actual detection system that is employed to measure individual compounds can be the same as in a total measurement. The flame ionisation detector (FID) is typical of this. Thus, a degree of overlap does exist in defining these two categories, and so it is unwise to consider this definition as exclusive. This article, however, will discuss one of the newer, latter techniques that have their origins firmly based in laboratory analysis, and which have been transferred to the industrial environment: Fourier transform infrared (FTIR) spectrometry.
All polyatomic gas molecules absorb infrared radiation at characteristic frequencies (wavelengths), with absorption intensity dependent on gas concentration. It is these properties which are used by all infrared analysers to give both qualitative and quantitative measurement of gas samples; typically VOCs, CO, CO2, NO, NO2, NOx, HCl, SO2.
While most standard infrared analysers used in industry are of the non-dispersive type, i.e. they measure at discrete wavelengths, the FTIR analyser measures a more complete part of the spectrum and uses multi-component analysis software to separate and identify the individual components. Originally these instruments were large, delicate and not easy to transport, and the last place you would choose to use them would be on a hot, wet, dirty gas stream. Today, the instruments have been made much more reliable, robust and transportable, providing an alternative to other multi-component techniques such as GC, MS or non-dispersive IR.
It is beyond the scope of this article to go into the details of construction of the various forms of analyser. In principle, the gas sample is passed through a sample cell (made of PTFE-coated aluminium or stainless steel) between gold-coated mirrors to give a long pathlength, up to 8m or more, to give a resolution down to sub-ppm levels.
Because a complete spectrum of the sample gas is given, it is possible, using multi-component analysis techniques, to identify the species present and, in the case of organic molecules, separate and quantify the concentration of individual molecules.
The sample is scanned using an interferometer which acts to separate the wavelengths of absorption onto a detector - a thermoelectrically-cooled device which gives low-ppm detection limits in 30 seconds or less.
As with all instrumentation, there is no universal solution to all problems. Low-resolution FTIR instruments can be made simple and rugged, and high signal-to-noise ratio can be obtained without the use of liquid-nitrogen-cooled detectors. Due to the lower absorbency values and lower noise levels obtained with the low-resolution technique, an increased dynamic range is given compared with that from a high-resolution instrument.
The major limitations of the low-resolution instrument are that there is a level of spectral overlap, and deviations from Beer's Law are seen due to absorbance non-linearity. These effects can be compensated for mathematically, but are cited as a drawback to the technique. In contrast, high-resolution instruments can compensate readily for non-linearity, and spectral overlap is avoided. However, the instruments tend to be bulkier, less robust and more difficult to set up.
The great attraction of FTIR is the ability simultaneously to measure a number of components (often >30) on a continuous basis and to analyse for unknown components. Typical applications include the following:
Maintenance costs can be much higher, and require the use of specialist assistance from the manufacturer.